YT YouTube2Text

YouTube2Text → Transcripts → Line coding, modulation, framing and multiplexing

Line coding, modulation, framing and multiplexing — Transcript

by Gunnar Karlsson · 3,152 words · 473 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00EP 1100 data communication and computer
0:04networks some illustrations in this
0:07material are collected from the book by
0:08foron data Communications and networking
0:11published by
0:14mcrill in the previous Parts I've shown
0:17examples of transmission medium and we
0:19also looked at the ultimate limits of
0:23data communication that you can get from
0:24a given medium the Shannon
0:27capacity now I will show you some actual
0:30methods of transmitting data over
0:33transmission
0:35medium we have two examples called
0:38coding and modulation if you have a low
0:40pass Channel which permits frequencies
0:43from zero up to some maximum limit you
0:47can use something called line coding to
0:49carry the data or it's also referred to
0:52as baseband
0:53modulation it's basically a direct
0:56current or a continuous light signal
0:59that is being modulated to carry the
1:02data if you instead have a band pass
1:05Channel then you need a carrier signal
1:08that lies somewhere in the pass band and
1:11you modulate that in order to carry the
1:13data these are the two principles that I
1:15will show
1:17you the line coding of Bas band
1:19modulation is used to turn data into
1:22transmission signal it uses discrete
1:24amplitude levels and the transitions
1:27between the levels occur at discrete
1:30times so it's a clocking where each
1:33clock interval corresponds to one or a
1:36fixed number of
1:38bits the signal is constant between
1:41these transmission times as shown in the
1:42figure up to the right you can see that
1:45a zero level corresponds to a zero in
1:48the data and a high level corresponds to
1:51a one in the data and you see that the
1:54signal remains low or zero and or high
1:58but has no intermediate values and
2:00there's a sharp transition between the
2:02low and the high
2:04levels the effect is that if you create
2:07such a signal and look at the forer
2:10Spectrum of the signal it has an
2:11infinite bandwidth while it's being
2:14transmitted through the low pass Channel
2:16some of the higher frequencies will be
2:18attenuated and you will only receive the
2:21the lower frequencies of of the signal
2:23Spectrum this means that the signal
2:25that's being received is not as sharp in
2:28the transitions as was
2:32Cent here's an actual case of a line
2:35coding it's called unipolar encoding and
2:37it's example that was used in the
2:39previous
2:40picture this signal contains a direct
2:44current component a Sero frequency
2:46component you can imagine if you only
2:50transmit an infinite string of ones then
2:52you will have a fixed voltage and you a
2:55current will flow through the
2:58cable the average level level of the
3:00signal will thus depend on the number of
3:03ones that you have in the signal but it
3:04will always be above zero this type of
3:08coding is only possible if you have a
3:10cable that permits a direct current to
3:12flow through the
3:14cable important that the receiver is
3:16able to clock the signal so so it should
3:19know how long the bit intervals are as
3:22seen from the
3:24transmitter in this encoding if you
3:27transmit a long stretch of only zeros or
3:30only ones you can imagine that the
3:33receiver will not have any transitions
3:36to indicate the edges of a an interval
3:40therefore the receiver might lose
3:42synchronization with the sender so we
3:45would like to have a signal with a lot
3:47of transitions because the transitions
3:49Mark the beginning and the end of the
3:51bit intervals so the receiver has some
3:53information to to synchronize its clock
3:56to here is another example of line
3:58coning this is called non- Return To
4:00Zero NR Z the picture shows two ways of
4:04encoding the data in the upper figure
4:08the data is encoded by the level of the
4:10signal so a positive level is represents
4:13a zero and a negative level represents a
4:16one if zeros and ones are equally likely
4:19in the data then you will not have a DC
4:21component in this signal in the lower
4:24figure we show the same non- return the
4:27zero code but it encodes the data in a
4:29different way it has a transition if the
4:32next bit is a one the advantage by
4:35coding through the transitions is that
4:38the signal becomes less sensitive to
4:41attenuation because it's not the level
4:43that's been detected it's the transition
4:45between the levels that is detected so
4:47you see here zero remains a high level
4:51and then the next bit is one so there's
4:53a transition and then after the one
4:56there's two zeros so there's no
4:57transitions then there comes a one so
4:59there is a transition actually the three
5:02ones in a row so there three transitions
5:05each encoding one
5:06bit synchronization is still a problem
5:09with NR Z because we can still have very
5:13long stretches of zeros and ones so that
5:16there are no transitions at all and
5:18therefore the receiver will not have
5:20information to synchronize its clock
5:23to so here is another line code called
5:26Return To Zero which always have a
5:28transition in order to make sure that
5:30there's clocking information to the
5:32receiver we use two signal elements per
5:35bit interval so one bit is encoded as
5:38two
5:39elements the the first element is either
5:43positive or negative and the second
5:45element is always zero which means that
5:48in every interval there's always a
5:50transition going from positive to zero
5:53or going from negative to zero so in
5:55terms of DC component it's similar to NR
5:59Z but it has a better clocking
6:03frequency however the board rate is now
6:05twice the bit rate because we have two
6:07signal elements for each bit transmitted
6:10if we willing to have a higher board
6:12rate we can also have a Manchester code
6:14Advantage compared to to The Return To
6:17Zero is that here we only use two
6:20different levels we don't have a zero
6:22level we have a positive and a negative
6:24level and there's a a transition so that
6:27a zero is represented from from positive
6:29to negative and a wers represented from
6:33negative to to positive the advantage
6:35with the Manchester code is that it
6:37there is a transition in every bit
6:39interval so that the receiver can
6:42synchronize its clock to the sender
6:44clock another possibility to make sure
6:47that you have good clocking information
6:49is to use a block code this code is
6:51applied to the data it transforms the
6:54data before the line
6:56coding here we show show something
6:58called 4B 5B so four bits in five bits
7:03out so you take the data you group it
7:06into four bit patterns There are 16
7:09combinations of those and you map those
7:11into five bit patterns so we have 32
7:15different outgoing
7:17patterns and we choose the outgoing
7:20patterns so that we have a certain
7:22number of ones and
7:25zeros so even if we encode the data by a
7:29nonr Return To Zero code we are sure
7:31that there will be Transitions and we
7:33have clocking
7:34information uh to the receiver there are
7:38also other variants such as 8 bit 10bit
7:41codes here you increase the data rate by
7:43going to four to five bits or from 8 to
7:4610 bits but on the other hand you can
7:49use a simple line code such as non-
7:50Return To Zero which has a board rate
7:52which is equivalent to the data rate
7:55rather than using Return To Zero or
7:58Manchester which has a board rate which
8:00is twice the data
8:02rate
8:04modulation modulation is the technique
8:07we use when we have a bypass Channel we
8:09take a sinusoidal wave we position it
8:12well in the pass band and then we
8:15modulate one or several of the
8:18parameters of the sine wave you see here
8:20the function s of T is a sinusoidal it
8:23has an amplitude a it has a frequency F
8:27and it has a phase fi
8:29so these are the three parameters that
8:32we can change to represent data a bit
8:35interval will be some duration of the
8:38sinusoidal where it remains constant
8:41before we change it to represent another
8:43data bit here we show amplitude shift
8:46keying ask it uses two amplitudes a
8:49lower and a higher one you see in the
8:51first interval that a zero is
8:53represented by a low amplitude and then
8:56in the second interval a one is
8:58represented by a higher amplitude this
9:00type of encoding is sensitive to um
9:03attenuation because the amplitudes will
9:07tend to zero and of course it's also
9:10subjected to noise that will be added to
9:12the signal so that can be difficult to
9:14detect whether the received signal is
9:16low or high in amplitude then we can use
9:19the frequency we show uh the Zer bit
9:22being coded by a lower frequency and the
9:24bit value one is represented by a higher
9:27frequency the signal strength is the
9:29same throughout the signal so that means
9:32that all intervals are affected equally
9:34by attenuation and the attenuation of
9:36course does not affect the frequency so
9:38the limitation here is the bandwidth of
9:41the pass band of the
9:43medium a third possibility is to use the
9:46phase so you see the zero bit is
9:49represented by one phase and the one bit
9:52is represented here by phase which is
9:55180° shifted from the zero bit the
9:58signal level remains constant for the
10:00signal and the detector has to detect
10:03the phase changes from one bit to to the
10:06next you can combine these different
10:09formats and here is an example called
10:11quadrature amplitude modulation which
10:13uses two amplitude values and four face
10:16values so that gives you uh two bits for
10:20the face values and one bit for the
10:22amplitude values so in total each signal
10:24element can encode three bits so I now
10:28shown how you can represent data as
10:31signals that travel through a medium now
10:34we go over some data transmission modes
10:37you can transmit data in parallel or
10:40serially if you transmit it serially you
10:42can have an as synchronous and
10:44synchronous mode the parallel
10:46transmission is most common inside
10:49computers a computer bus or any data
10:53path on a printed circuit board are
10:55usually carrying the data in parallel so
10:58here we show that that eight bits are
11:00transferred in parallel at any one time
11:02there could be additional conductors
11:05that carry for instance clocking
11:06information as well this is also what
11:08you have in parallel cables for instance
11:10to connect a printer serial transmission
11:13takes the data which might be structured
11:15in data words or bytes and then it sends
11:18them serly one bit at a time it's
11:21important that the sender and the
11:22receiver agree on the order of the the
11:24serialization for instance if you have a
11:27bite should the most significant bit be
11:29sent first or should the least
11:30significant bit be sent first either
11:33order will work but the sender receiver
11:36have to agree on the order here you need
11:38some other clocking information external
11:40clocks of course but uh the type of line
11:43coding uh that that I showed could be
11:45used here so that the signal itself
11:48carries enough timing information for
11:50the receiver to synchronize its clock to
11:52the next aspect of data communication is
11:54framing of messages a sender will not be
11:58active continuously for all times it
12:01will have data to send and then when the
12:03data has been sent it may remain idle
12:06for a long period of time this has to be
12:08communicated to the receiver so that the
12:11receiver knows that this end of message
12:13and then whatever signal it receives
12:15will be noise generated in the medium
12:18and then later on when a new message
12:20comes the receiver has to be able to
12:22detect that there's a new message
12:23starting and it should now detect the
12:26signal and and translate the signal into
12:29data we call this framing so it marks
12:32start and stop of
12:34messages it's some Control Data that's
12:37added for each message so the data
12:39that's generated from a higher layer
12:42ultimately from an application will now
12:44be framed by by some predefined bit
12:47sequence at the beginning and end of the
12:50message the framing also allows
12:52multiplexing of messages from multiple
12:55sources if each source is only active a
12:58small small percentage of time then
13:01several sources could share a channel
13:03for instance a radio channel there will
13:05therefore be frames coming separately
13:07from each Source towards the receiver
13:10could also be that a frame is shared by
13:12multiple sources that are collocated so
13:15they they can put the data in different
13:17parts of one and the same
13:19frame
13:22multiplexing multiplexing is the concept
13:24of sharing a link here we show a simple
13:27example where the four possible inputs
13:30on the left side to the multiplexer
13:33there's one link carrying a bitstream or
13:37frames of bits but it's divided into
13:39four channels and then the multiplex
13:42will separate out these channels and
13:44deliver them so we have emulated here a
13:47situation where we have four different
13:49links outside the max but they're
13:52implemented as one link um in one
13:55transmission medium this allows multiple
13:58sender receiv receiver pairs to share a
14:00link and to share the resources of that
14:03link because it could be that the
14:04resource of the medium is so high that
14:07not any single sender would generate as
14:09much data as the medium can can carry
14:12for instance um fiber optical cable
14:16there are several ways of sharing the
14:17resource of the transmission medium one
14:21is that we look at the bandwidth of the
14:23transmission medium and we divide that
14:25into different frequency channels and
14:28let each send a receiver pair have its
14:31own reserved uh frequency range another
14:34thing is that we modulate or or line
14:37code for the entire bandwidth of the
14:40medium and then we divide up that
14:43bitstream into a Time units and each
14:47time unit correspond to the slot that a
14:50sender will have available to send its
14:52data in the first case we talk about
14:54analog multiplexing and this method of
14:58dividing up the bandwidth into frequency
15:00regions called frequency division
15:02multiplexing it uses a modulation for
15:04each frequency Channel we have a carrier
15:06frequency that is positioned in the
15:08channel that has been allocated this is
15:11the method that's been used for
15:12broadcast radio and television where for
15:15instance different radio channels uh
15:18correspond to a frequency range that
15:20they have the center frequency which is
15:22used by the carrier frequency and the
15:23bandwidth around it that they have to
15:26remain
15:27within the concept is also used in
15:29Optical networks where it's referred to
15:31as wavelength division
15:33multiplexing it's the same concept as
15:35fdm and here it corresponds to sending
15:38the data with different wavelength
15:40channels or different colors of the
15:42light that's being
15:45transmitted and then we have digital
15:47multiplexing where we let the sender
15:49generate a data rate for the entire
15:52bandwidth of the medium then we divide
15:55the times into slots and uh allocate
16:00slots to the different senders so the de
16:03multiplexer will then know that in each
16:05frame the first slot belongs to channel
16:08one the second slot belongs to channel
16:11two and it will separate out the slots
16:13and deliver the data to the receivers of
16:15those channels this is called
16:18asynchronous time division multiplexing
16:20we can also have an asynchronous time
16:22division multiplexing where each Source
16:25gets its own frame that's sent into the
16:27medium the frame will then have an
16:30address to indicate which channel that
16:33it belongs to so that the demultiplexer
16:36can take the frame and deliver it to the
16:38right receiver here is an example of
16:41synchronous time division Multiplex info
16:43telepan so as we heard in the PCM part
16:47the 4 KZ of voice spectrum is PCM coded
16:51resulting in 64 kilobits per second of
16:54data such streams were multiplexed one
16:57BTE per input and put into frames in the
17:01United States there were 24 channels in
17:03Europe and many other parts of the world
17:05there were 30 channels resulting in a
17:07total bit rate of 2 megabits per second
17:10and then there's whole hierarchy of
17:12these E1 or T1 channels being
17:15multiplexed together to higher data
17:17rates up to hundreds of megabits per
17:19second here's another example the
17:22digital subscriber
17:24Loop this is a high-speed digital access
17:27that uses the old telephone wires that
17:30go into most homes it uses the actual
17:33bandwidth in the twisted pair cable um
17:36in the subscriber access
17:38lines even though an analog voice call
17:41was only 4 KZ in bandwidth the cables
17:45permit up to over megahertz in in
17:49bandwidth the bandwidth of the medium is
17:51subject to strict physical imitations it
17:54has a high attenuation so you can only
17:56go short distances and it requires a
17:58Advanced signaling this is one way that
18:01the uh subscriber Loop is being used it
18:05takes the full spectrum of the medium
18:06and divides it into frequency bands
18:08there is one band reserved for
18:10traditional analog voice
18:13communication and then here you see
18:15there are bands that are not being used
18:17and then other bands that used for
18:19Upstream because this particular mode is
18:21called asymmetrical
18:23DSL then there are other channels more
18:26channels used for Downstream data so
18:28that that's why it's asymmetrical you
18:29get more data in the downstream
18:31Direction the bandwidth of the channels
18:33is 4 khz and then there's a adaptation
18:37so that the data rate per channel is
18:40adapted to the conditions of that
18:43channel there could even be that an
18:45individual channel is not used at all if
18:47it's very poor quality this is used for
18:49residential a Broadband access for
18:52data in the Upstream Direction typically
18:54up to 1 megabit per second and in the
18:57downstream Direction it megabits per
18:59second um but nowadays higher 24
19:03megabits and up to 100 megabits per
19:05second depending on the quality of the
19:07cable this shows how it's accomplished
19:10there's a voice Channel which is just
19:13transparent and then there's a Serial to
19:16parallel converter here so the data is
19:18sent in parallel over the different
19:20channels and then there's an advanced
19:22modulation quat to amplitude modulation
19:25which has 15 bits per board for each
19:28Channel
19:29and you see the same modulation being
19:31used for the downstream
19:33channels so each Channel being 4 KZ and
19:37there's 15 bits per board it means that
19:40you can carry 60 kilobits per second per
19:43channel so in summary for this teaching
19:45module I've showed transmission media
19:47how you can see look at the link budget
19:49and the capacity limitations of it we
19:52have in this part looked at uh
19:54transmission of digital information by
19:56line coding and digital modulation
19:59I've briefly mentioned synchronization
20:01and multiplexing and shown examples of
20:04multiplexing such as PCM of digital
20:07telepan or synchronous digital
20:09subscriber Loop

About this transcript

This page contains the full transcript of Line coding, modulation, framing and multiplexing by Gunnar Karlsson, generated from the public captions YouTube serves with the video. The transcript has 3,152 words across 473 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.